DEVELOPING GUIDANCE ON MONITORING AND PROJECT … · developing guidance on monitoring and project boundaries for greenhouse gas projects information paper. com/env/epoc/iea/slt(2002)2

Organisation for Economic Co-operation and Development 2002International Energy AgencyOrganisation de Coopération et de Développement EconomiquesAgence internationale de l'énergie

COM/ENV/EPOC/IEA/SLT(2002)2

OECD ENVIRONMENT DIRECTORATEAND

INTERNATIONAL ENERGY AGENCY

DEVELOPING GUIDANCE ON MONITORINGAND PROJECT BOUNDARIES

FOR GREENHOUSE GAS PROJECTS

INFORMATION PAPER


2

FOREWORD

This document was prepared in May 2002 by the OECD Secretariat at the request of the Annex I ExpertGroup on the United Nations Framework Convention on Climate Change. The Annex I Expert Groupoversees development of analytical papers for the purpose of providing useful and timely input to theclimate change negotiations. These papers may also be useful to national policy makers and otherdecision-makers. In a collaborative effort, authors work with the Annex I Expert Group to develop thesepapers. However, the papers do not necessarily represent the views of the OECD or the IEA, nor are theyintended to prejudge the views of countries participating in the Annex I Expert Group. Rather, they areSecretariat information papers intended to inform Member countries, as well as the UNFCCC audience.

The Annex I Parties or countries referred to in this document refer to those listed in Annex I to theUNFCCC (as amended at the 3rd Conference of the Parties in December 1997): Australia, Austria,Belarus, Belgium, Bulgaria, Canada, Croatia, Czech Republic, Denmark, the European Union, Estonia,Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Latvia, Liechtenstein,Lithuania, Luxembourg, Monaco, Netherlands, New Zealand, Norway, Poland, Portugal, Romania,Russian Federation, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey, Ukraine, United Kingdom ofGreat Britain and Northern Ireland, and United States of America. Where this document refers to“countries” or “governments” it is also intended to include “regional economic organisations”, ifappropriate.

ACKNOWLEDGEMENTS

This paper was prepared by Jane Ellis (OECD). The author would like to thank Jan Corfee-Morlot andStéphane Willems (OECD), Martina Bosi and Jonathan Pershing (IEA) for the information, comments andideas they provided. The author is also grateful to Jan-Willem Bode, Johannes Heister, Anke Herold,Klaus Radunsky, Lasse Ringius and Einar Telnes for their advice and comments.

Questions and comments should be sent to:

Ms. Jane EllisAdministratorOECD Environment Directorate2, rue André Pascal75016 Paris CedexFranceEmail: [email protected]: +33 1 45 24 78 76

OECD and IEA information papers for the Annex I Expert Group on the UNFCCC can be downloadedfrom: http://www.oecd.org/env/cc/


3

TABLE OF CONTENTS

EXECUTIVE SUMMARY...........................................................................................................................5

1. INTRODUCTION..................................................................................................................................7

1.1 Role of monitoring in a GHG mitigation project..............................................................................8

2. KEY QUESTIONS.................................................................................................................................9

2.1 Should monitoring guidance allow choices in monitoring methods? ...............................................92.2 How can possible cost/accuracy tradeoffs be accommodated? ......................................................102.3 Project boundaries: what flexibility should project developers have?............................................13

2.3.1 Options to define “under the control” of the project participants .............................................142.3.2 Options to define “significant” emission or enhancement sources ...........................................152.3.3 Options to define “reasonably attributable” to the project ........................................................16

2.4 Should guidance on project boundaries be the same for emission reduction and sink enhancementprojects?.....................................................................................................................................................172.5 How can the monitoring burden be reduced? .................................................................................17

3. OTHER AREAS IN WHICH GUIDANCE MAY BE USEFUL .....................................................20

3.1 What a monitoring plan should contain..........................................................................................203.2 Frequency of monitoring, reporting, and verifying information.....................................................21

3.2.1 Monitoring and reporting frequency .........................................................................................213.2.2 Verification frequency...............................................................................................................213.2.3 Frequency/cost trade-offs..........................................................................................................22

3.3 Areas for detailed/sectoral monitoring guidance ............................................................................223.3.1 Specific guidance recommendations.........................................................................................243.3.2 Deciding which aspects of a project to monitor ........................................................................25

4. LESSONS LEARNED FROM EXPERIENCE IN MONITORING/VERIFYING PROJECTPERFORMANCE .......................................................................................................................................26

4.1 Project boundaries ..........................................................................................................................264.2 Frequency of monitoring ................................................................................................................31

5. CONCLUSIONS ..................................................................................................................................34

6. ANNEX A: MARRAKECH ACCORD PROVISIONS ON THE PROJECT CYCLE ANDPROJECT MONITORING........................................................................................................................38

6.1 CDM project cycle..........................................................................................................................386.2 Excerpts from Decision 17/CP.7 ....................................................................................................38

7. REFERENCES.....................................................................................................................................41

8. GLOSSARY..........................................................................................................................................45


4

LIST OF TABLES

TABLE 1: GENERAL RECOMMENDATIONS ON WHICH IMPACTS TO INCLUDE IN THE PROJECT BOUNDARY ...27TABLE 2: COMPARING PROJECT BOUNDARY RECOMMENDATIONS (ELECTRICITY/HEAT SUPPLY)................28TABLE 3: PROJECT BOUNDARY ASSUMPTIONS IN SELECTED CLIMATE MITIGATION PROJECTS

(ELECTRICITY/HEAT) .............................................................................................................................29TABLE 4: COMPARING PROJECT BOUNDARY RECOMMENDATIONS (REFORESTATION PROJECT) ..................30TABLE 5: PROJECT BOUNDARY ASSUMPTIONS IN SELECTED CLIMATE MITIGATION PROJECTS

(REFORESTATION)..................................................................................................................................31TABLE 6: SELECTED RECOMMENDATIONS ON FREQUENCY OF MONITORING ACTIVITIES FOR DIFFERENT

CARBON POOLS IN LULUCF PROJECTS.................................................................................................32TABLE 7: MONITORING FREQUENCY AND RESPONSIBILITIES FOR QUERCUS AIJ PROJECT ..........................33TABLE 8: IMPLICATIONS OF POSSIBLE ACTIONS TO REDUCE MONITORING COSTS ........................................37

LIST OF FIGURES

FIGURE 1: BASELINES AND MONITORING FOR AN EMISSION REDUCTION PROJECT.........................................7FIGURE 2: INFLUENCE OF NUMBER OF SAMPLE PLOTS ON THE MONITORING COST AND PRECISION LEVEL

FOR THE NOEL KEMPFF PROJECT...........................................................................................................12FIGURE 3: ABOVE AND BELOW-GROUND CARBON STOCKS IN DIFFERENT LAND TYPES ..............................15FIGURE 4: EXAMPLE OF A DECISION TREE DETERMINATION OF PROJECT BOUNDARIES (ELECTRICITY OR

HEAT GENERATION PROJECT) ................................................................................................................24


5

Executive Summary

Comparing the greenhouse gas (GHG) emissions or uptake of a project to its baseline allows the mitigationeffect of a project to be calculated. This mitigation effect corresponds to the maximum number of creditsthat a Joint Implementation (JI) or Clean Development Mechanism (CDM) project can generate.

The GHG emissions or uptake of a project is determined by monitoring its performance over time. TheMarrakech Accords indicate that a project monitoring plan needs to be established as part of the JI/CDMproject design document. Monitoring may also be needed at other times or for other purposes. Forexample, monitoring can be used to establish project baselines or to determine whether or not a project’sinitial crediting period can appropriately be extended.

This paper focuses on the monitoring needed to quantify the impact of a GHG mitigation project and ondefining the boundaries of what needs to be monitored. Drawing on lessons from project monitoringexperience and guidance to date, it identifies key questions whose answers set the ground rules for specificmonitoring guidance for project types or categories. It also identifies other, more detailed, areas in whichmonitoring guidance would be useful.

Developing guidelines for GHG mitigation projects on baselines and monitoring are related and should bedone in parallel. Monitoring guidance can encompass broad recommendations, such as what constitutes a“significant” emissions source, as well as detailed technical recommendations, such as where to locatemeters or how to read them.

Some key questions need to be answered before detailed monitoring guidance can be set up:

• Should project developers have a choice of different possible monitoring methods?

• What flexibility should project developers have in defining the project boundary?

• How should any general cost/accuracy tradeoffs be accommodated?

The answers will influence the extent to which implementing monitoring guidance promotes consistency,transparency and predictability between different projects. The answers will also affect the time and costassociated with verifying a project’s emission reductions.

This paper recommends allowing a choice of monitoring methods. This will give project developers someflexibility to use less complex, accurate or costly monitoring methods for less important emission sources.For some project types, standardised measurement and monitoring protocols may have already been set up.

Defining project boundaries is important – for both baselines and monitoring - as it affects both thenumbers of credits that a project can generate and the potential significance of leakage from a project.However, the Marrakech Accords’ definition of project boundary is not precise enough to ensure thatdifferent project developers will reach consistent and predictable decisions. Defining wide projectboundaries and using simple monitoring methods for less important emission sources may be anappropriate balance between reducing the risk of leakage and limiting monitoring costs. Developingdecision trees to guide project developers as to what to include in a project boundary, and how rigorousemission methodologies should be for each source, may also be appropriate.


6

There are costs associated with developing and carrying out a project monitoring plan. The costs associatedwith developing a monitoring plan are occurred up-front, and can be high (e.g. $20,000 for “first-of-a-kind” CERUPT and PCF projects). Some costs associated with monitoring a project’s performance arelikely to be incurred whether or not a project has been set up as a greenhouse gas mitigation project.However, some items that need to be monitored in order to assess the GHG benefits of a project (e.g.increase in underground biomass) would not normally be monitored unless the project was seeking to gainemissions credits. Thus, the importance of the JI/CDM-component of on-going monitoring costs alsovaries by project type. Project developers can be expected to prefer projects that are easy to monitor inorder to keep total monitoring costs below a certain level, e.g. 8% of the total costs of a project.

The on-going costs of monitoring a project’s performance can vary significantly by project type. Ingeneral, monitoring costs will increase with the number and variety of sources that need to be monitored;frequency of monitoring1 and reporting; accuracy/confidence level that monitoring should achieve; and thecomplexity of monitoring methods. Ideally, project monitoring should provide an accurate picture (or, atthe least, not a consistent over-estimation) of a project’s GHG impacts. However, while monitoring can intheory be a more-or-less exact exercise, developing emission baselines is not. It may therefore beappropriate to consider possible trade-offs when developing monitoring guidance. This paper recommendsthat projects using less accurate monitoring methods, or lower monitoring/reporting frequencies for majoremission sources should generate fewer credits than similar projects applying more rigorous or regularmonitoring methods.

Simplifying the monitoring process to reduce monitoring costs may be particularly important for small-scale CDM projects that, by their very nature, are less able to bear high transaction costs. Thesesimplifications could include the use of sample populations, reducing the frequency ofmonitoring/verification and calculating project benefits by using easy-to-monitor data, default emissionfactors and narrow project boundaries. For the most environmentally-friendly projects, such as electricitygeneration from wind or solar power, highly simplified monitoring could be limited to ensuring that theequipment installed continues to operate.

Once the ground rules for monitoring guidance have been defined, more specific guidelines are needed tooutline e.g. exactly what should be monitored, when, how, and by whom. This specific guidance canbenefit greatly from input and/or lessons from guidance developed elsewhere, e.g. for developingemissions inventories, monitoring GHG mitigation projects, or from national or international monitoringstandards in particular sectors.

The Marrakech Accords indicate that a project’s monitoring plan should include a description andjustification of the project boundary, identification of data that needs to be collected, qualityassurance/control procedures and a list of potential sources of leakage. These recommendations may needto be elaborated and expanded. For example, the monitoring plan should also indicate how often differentparameters should be monitored and reported; who is responsible for monitoring and recording thedifferent information; how the monitored information should be recorded, and what information is likely tobe needed in order to verify credits. An outline of any training that project operators will need to carry outthe GHG-related monitoring and reporting activities in the monitoring plan may also be needed.

Examining actual or recommended project boundaries for GHG mitigation projects shows that differentguidance on what to include in the project boundary can differ markedly (for a particular project type). Theneed for transparency and consistency between different projects reinforces the value of detailed guidanceon project boundaries.

1 Unless continuous emissions monitoring is used.


7

1. Introduction

One of the reasons for setting up a Joint Implementation (JI) or Clean Development Mechanism (CDM)project is to generate emission credits. These credits can be used by Annex I Parties to help meet theiremission commitments under the Kyoto Protocol (KP). Domestic greenhouse gas (GHG) mitigationprojects may also be set up and need to be monitored in the context of e.g. voluntary programmes to reduceemissions in particular sectors.

In order to determine the mitigation effects of such projects the emissions or uptake level of the projectneeds to be monitored and then compared to that of the project’s baseline (Figure 1). For JI/CDM projects,the extent that emissions have been reduced below the baseline, or that removals have been increasedabove the baseline, corresponds to the maximum number of credits a project can generate2.

Figure 1: Baselines and monitoring for an emission reduction project

Monitoring is defined here as the measurement or estimation of the greenhouse gas (GHG) impacts of amitigation project.

The Marrakech Accords indicate that monitoring project performance is an essential part of the JI/CDMproject cycle. While the Accords do lay out the role of some actors in monitoring, almost no otherguidance is provided. However, if project developers and operators knew what the mandatory “monitoringplan” should consist of, what aspects of the project need to be monitored, how they should be monitored,when and by whom3 it could reduce the time and costs associated with project development. Applying suchmonitoring guidance could therefore increase transparency, predictability and comparability of creditgeneration from projects. It could also increase the ease and reduce the costs of verifying emissionreductions.

2 In practice, the number of credits obtained by project developers may not equal the number of credits generated bythe project. Some CERs will be deducted from CDM projects in order to cover the administrative expenses and assistin meeting the costs of adaptation. In addition, some host countries may negotiate to keep a proportion of CERsand/or ERUs themselves.3 Similar information will also be needed to quantify the effects of any project-based reductions within a nationalcredit programme.

Time (years)7 14 21

Em

issi

ons

(to

talG

HG

orG

HG

/uni

tout

put)

Number of creditsgenerated by project(calculated)

Baseline(hypothetical)

Project emissions(monitored)


8

While detailed guidance on what and how to monitor will vary by project category, some generic guidanceon other aspects of project boundaries and monitoring should be drawn up. This paper seeks to clarifysome of the relevant key issues. Drawing on lessons from project monitoring experience and guidance, itidentifies key questions that need to be answered before any generic (or specific) project guidance onproject monitoring and boundaries can be agreed, and makes recommendations on how to answer suchquestions. It also makes some recommendations as to what could be considered for inclusion in anymonitoring guidance to be developed by the CDM’s Executive Board.

1.1 Role of monitoring in a GHG mitigation project

There are several reasons to monitor particular aspects of JI/CDM projects, and the context in whichprojects are (or will be) operating. These include to:

• Monitor the GHG emissions/uptake of the project itself. This information is needed to establish thenumber of credits generated by the project (by comparing to the baseline);

• Pre-project monitoring in order to estimate an appropriate baseline for the project. This may or maynot be needed depending on how the baseline for a project is being set up;

• Assess the wider environmental or other impacts of the project (e.g. the importance of carbon leakageand spillover). This information is needed to assess whether or not the number of credits generated bythe project represents the actual mitigation effect of the project;

• Ensure (for forestry projects) that the GHG benefits of the project are not reversed, i.e. that if“permanent” credits are generated by LULUCF projects, the sequestration is also permanent4;

• Draw lessons ex post from projects already implemented so as to provide feedback and insights on howwell baselines and other rules in the project cycle are working;

• Determine whether or not the project’s initial baseline can appropriately be extended for a second orthird crediting period.

Thus, monitoring will need to be carried out at different points during the development and operation of aGHG mitigation project. Most obviously, monitoring will need to take place during the lifetime of a projectin order to quantify the GHG impact of the project (see text box). Such monitoring is likely to need to becarried out on a regular basis, e.g. annually, in order to obtain a regular flow of emission credits.Monitoring outside the project boundary during the project’s lifetime may also be necessary in order toassess the importance of leakage. Over time the findings of such an assessment could influence projectdesign and any revisions/updates on rules for eligibility and baselines.

What information needs to be monitored, and who should be responsible for monitoring will differdepending on the reason for which monitoring is taking place. This paper focuses on the monitoringneeded to quantify the GHG impact of the project itself. This paper also explores the project boundaryissue in some depth.

4 This aspect of monitoring forestry projects may or may not be needed depending on the crediting regime used. SeeEllis 2001 for a more detailed discussion.


9

2. Key questions

The development of monitoring guidelines for JI and CDM projects is related to the development ofrecommendations on baselines and additionality. There are also strong links between monitoring guidance,and guidance on reporting and verification. Development of guidance for baselines, monitoring andreporting should be done in parallel. As for guidance on baseline-setting, monitoring guidance can rangefrom very broad guidance, such as what constitutes a “significant” emissions source to detailed technicalrecommendations, such as where to locate meters, or how to read them.

Answers to some key questions will define the ground rules needed before detailed monitoring guidancecan be set up. These questions centre around the level of variation in monitoring the performance ofdifferent projects that is compatible Marrakech Accords’ indication that general monitoring guidance willbe set up in order to “promote consistency, transparency and predictability”. Answering these questionscan help set the ground rules for any subsequent detailed monitoring guidance, and could significantlyinfluence project monitoring costs. The main questions are:• Should monitoring guidance allow choices in monitoring methods?

• How could any general cost/accuracy tradeoffs be accommodated?

• What flexibility should project developers have in defining the project boundary?

• Should guidance on project boundaries follow the same principles for emission reduction and sinkenhancement projects?

2.1 Should monitoring guidance allow choices in monitoring methods?

The answer to this question is key to setting the ground rules for any detailed monitoring guidance. Manydifferent possible methods could appropriately be used to monitor emissions or enhancements by a project.For example, carbon uptake by standing biomass (trees) in forestry projects can be estimated from timberinventory techniques (on-site measurements), modelled, estimated via the use of allometric equations orestimated by using a combination of these methods. These different methods can result in differentestimates of a project’s emissions or enhancements, at different costs, and with different levels of accuracyand confidence.

Both the IPCC guidance on good practice in inventory calculation (IPCC 2000) and international guidancedeveloped to determine the results of energy efficiency projects (IPMVP 2000) allows for a choice ofmonitoring methodology. For example, the IPCC inventory guidance often includes two or three differentmethodologies to calculate a particular emission, with more detailed methods being recommended for“key” emission sources.

• Despite potential differences in the results, there may be a number of reasons to leave projectdevelopers flexibility in the choice of monitoring methods. The Marrakech Accords allow a choice ofbaseline methodology. For example, these methodologies do not rule out the use of “absolute”baselines, which would require different elements to be monitored than if “rate-based” baselines wereused. In order to ensure that baseline and monitoring guidance are compatible, a choice of monitoringmethods will be needed.


10

• The relative importance of different gases and sources can vary enormously, depending on thecharacteristics of the individual greenhouse gas mitigation project. It may be cost-effective to allowproject developers to use simplified methodologies to monitor sources that are not “key”.

• Different equipments and methods/techniques may be already in place in different countries.Recognising such differences and building on existing processes may be necessary.

Recommendations:

� Project developers should be allowed flexibility in the choice of monitoring methods theyuse. This will allow developers to use simplified methodologies to monitor minor sources.� Once a particular method is used it should be applied consistently throughout the projectlifetime.

2.2 How can possible cost/accuracy tradeoffs be accommodated?

Each part of the JI/CDM project cycle has associated transaction costs. Monitoring the performance andGHG impacts of a project can be an important component of a project's total transaction costs for someproject types.

Developing a monitoring plan may also be costly, particularly for a “first-of-a-kind” project. For example,the Prototype Carbon Fund (PCF 2000) estimates that developing a first-of-a-kind baseline, monitoring andverification protocol (MVP) takes about 4-5 weeks and costs approximately $40,000. However, they alsosuggest that subsequent MVPs for similar project types will be much quicker and cheaper to set up. Themonitoring and verification component of this accounts for approximately $20,000 (Heister, 2002). Costsfor developing a first-of-a-kind CERUPT project are estimated at $20,000 (CDM Susac 2002), althoughthese costs may halve for subsequent projects (Bode 2002). This level of transaction costs at a stage whenthe proposed project has not even been approved, is likely to inhibit project development, particularly forsmall projects. Establishing guidance on what should be included in a monitoring plan could help projectdevelopers significantly reduce the costs associated with setting up a monitoring plan.

The cost of monitoring aspects of a project that are relevant to generating emissions credits, and that wouldnot normally be monitored for non-JI/CDM projects will therefore need to be kept low (e.g. 1-8% of thetotal project cost) in order to not inhibit project development.

The Marrakech Accords make it clear that resources will need to be devoted to project monitoring from avery early stage in project development, and certainly before the project has been approved as a JI/CDMproject. The Accords also indicate that guidance on monitoring methodologies developed by the CDMExecutive Board needs to “provide an accurate measure of actual reductions in anthropogenic emissions asa result of the project activity, taking into account the need for consistency and cost-effectiveness”. Thisimplies that cost considerations may be taken into account when developing monitoring guidance. Forexample, it may be possible to choose to forego both the credits and the costs of monitoring of a particularemission source associated with an individual project5.

However, this appears to contradict text relating to project boundaries earlier in the same decision whichindicates that all “significant” sources that are “under the control” and “reasonably attributable” to aproject need to be monitored. Such text limits the choice that participants in JI and CDM emission

5 This may be most relevant for LULUCF projects, as some of the environmental benefit of these projects (e.g.increase in root or soil carbon) can be costly to quantify.


11

reduction projects have in deciding on any potential tradeoffs between the number of credits obtained andthe cost of obtaining those credits.

How onerous project monitoring is likely to be for CDM projects is not yet clear, as it will depend on thedefinition of project boundaries as well as on the methods recommended to monitor particular sources ofemissions or enhancements. Actual project monitoring costs can vary widely. For example, Ecosecurities(2000) estimate that transaction costs of JI projects will vary between $10,000 to $15,000/year. ThePrototype Carbon Fund indicate that monitoring and verification costs are likely to account forapproximately half of the total $200,000 - $400,000 transaction costs of the project (over the lifetime of theproject)6. The IPMVP, which outlines rigorous methods to assess energy savings of energy efficiencyprojects estimates that this assessment should cost no more than 10% of the annual average savings beingassessed. GEF energy efficiency projects (Birner 2001) set aside between 1-8% of a project’s overallbudget for monitoring and evaluation purposes. Theoretical case studies in Nabuurs et al. (1999) indicatethat the costs of monitoring Kyoto forests in JI countries may vary from $11-18/hectare.

The costs of monitoring project performance will vary by type of project and project design, but are likelyto increase with the:

• Number and variety of sources that need to be monitored;

• Frequency of monitoring and reporting;

• Accuracy/confidence level that monitoring should achieve; and the

• Complexity of monitoring methods.

The number and variety of sources that need to be monitored varies by project type. For example, there isonly one main location for emission reductions in projects that involve emission reductions in single, pointsources - such as large electricity-generation or industrial projects. Monitoring emissions on one site islikely to be easier and cheaper than monitoring projects that involve diffuse emissions reduction or uptake,such as in energy-efficiency or forestry projects (although monitoring these projects can also benefit fromeconomies of scale, or sampling). Similarly, monitoring only CO2 emissions from a project is likely to beeasier and cheaper than monitoring CO2, CH4 and N2O emissions.

Increased frequency of monitoring and reporting will also increase costs, although not necessarilysignificantly if what needs to be monitored for GHG purposes would have been monitored anyway. Forexample, grid-connected electricity generation projects are likely to keep track of their fuel consumptionand electricity generation or output whether or not they are a JI/CDM project. However, information whichmay be needed to calculate the effect of energy efficiency projects, such as the amount of energy used forlighting, in which rooms, and for how long, are unlikely to be collected as a matter of course (unless set upas part of the project design). Similarly, soil and root carbon would not normally be estimated for forestryactivities unless they were aiming to generate emission credits.

Any requirement to increase the accuracy or confidence level with which a project emissions orenhancement calculation is made could also increase the costs of project monitoring. This could meanincreased fixed costs, for example if different equipment needed to be installed. It could also increasevariable costs by increasing the numbers of forest sample plots or household appliance efficiency surveys

6 This includes monitoring some aspects of a project that may not need to be monitored for CDM projects. Forexample, PCF projects monitor baseline parameters (such as GHG intensity of electricity displaced by a project) andsustainable development indicators.


12

that need undertaking, or one-off costs (e.g. if project-specific emission factors needed to be measured,rather than using default factors).

Many methods could be used to monitor the carbon sequestration in forestry projects. These includemodelling, deriving estimates from equations and field measurements. These different techniques havedifferent associated costs, and their applicability and/or feasibility may vary with project size. However, inorder to verify the emission mitigation effect of a project, it is likely that field measurements (“ground-truthing”) will play an important role in monitoring and/or verifying changes in carbon stocks fromLULUCF projects. The number of sample plots and other field measurements used to monitor changes incarbon stocks has a large influence on the variable costs associated with project monitoring, although fixedcosts may be more important (Powell 1999). The number of sample plots used also has an influence on theaccuracy and confidence levels with which a project’s mitigation effect can be calculated (Figure 2).

Figure 2: Influence of number of sample plots on the monitoring cost and precision levelfor the Noel Kempff project

0

50

100

150

200

250

300

350

400

450

500

5% 10% 20% 30%

Precision level

Mo

nit

ori

ng

cost

s($

1000

s)

0

50

100

150

200

250

300

350

400

450

500

Nu

mb

ero

fsa

mp

lep

lots

Variable costsFixed costsNo. of sample plots

Source: adapted from Powell (1999) and Kadyszewski (2001)

However, preliminary indications are that the cost of sequestering carbon in forests is low. Brown et. al.(2001) indicate that the cost of measuring and monitoring carbon in (large) forestry projects would be lessthan $0.5 per ton of carbon for precision levels of less than +/- 10% of the mean with 90% confidence.This means that monitoring costs for (large) LULUCF projects could be somewhat higher than forenergy/industry projects and still be able to provide a low-cost sequestration potential.

The complexity of monitoring methods can also increase monitoring costs if it requires the use of moreexpensive equipment or of specially trained staff to carry out monitoring-related tasks, such as for forestryor energy efficiency projects.


13

In order to ensure that different projects produce CERs of equal “quality”, guidance may need to specifythe accuracy/confidence with which emission mitigation estimates need to be made. Alternatively, somesort of guidance or methodology could be set up that would allow a trade-off between rigour/accuracy ofmonitoring, and level of credit allocation. This type of approach has precedents, e.g. from the US EPA’sConservation Verification Protocols – where higher “credits” were given for energy savings estimatedthrough direct monitoring than through using e.g. standard equations. This approach has also beenrecommended as a potential way of dealing with the varying uncertainty, related to different methods, ofestimating the GHG mitigation effect of forestry projects (Hamburg 2000) and for energy projects (DNV2001b).

Recommendations:

� Keep “incremental” monitoring costs for larger GHG mitigation or sequestration projectsmanageable by ensuring that the level of precision/confidence required for project monitoringdata is comparable with that for baseline data.� Explore the feasibility setting up guidance that would allow a trade-off betweenrigour/accuracy of monitoring a project’s major emission impacts and the number of creditsallocated.

2.3 Project boundaries: what flexibility should project developers have?

The Marrakech Accords require that each JI/CDM project “shall” include a “monitoring plan” as part ofthe project design document (see Annex A). Defining the boundaries of what needs to be monitored is acrucial step in setting up a plan to monitor the project’s performance and is also needed when definingwhat the baseline for the project includes. In order to ensure that credits from projects credibly reflectGHG mitigation, parameters included in a project’s baseline need to be monitored, and vice versa. Thiswill ensure that the calculated emission benefit of the project compares emissions/uptake from the samesources.

How a project boundary is defined is important, because it influences the environmental credibility ofcredits generated by the project and the costs of monitoring (through the effect of project boundarydefinitions on the number of sources that need monitoring). For example, monitoring the emissions from azero-emitting project such as stand-alone renewable electricity generating projects is a trivial exercise, ifthe project boundary is drawn around the project site. However, if the project boundary has been drawn toalso include emissions associated with preparing the project site and transporting equipment to the projectsite, the complexity of monitoring operations increases.

The Marrakech Accords define project boundaries for emission reduction CDM projects as“encompass[ing] all anthropogenic emissions by sources of greenhouse gases under the control of theproject participants that are significant and reasonably attributable to the CDM project activity”7. Thisdefinition therefore sets up a three-staged approach to defining project boundaries. Only emission sourcesthat are significant and reasonably attributable and being “under the control” of the project participantsneed to be included in the project boundary. Thus, emission sources that are e.g. significant but not underthe control of the project participants are not included in the project boundary. For example, an electricitysupply project that increases electricity generation near centres of electricity demand may significantlyreduce losses due to electricity transmission and distribution (T&D). However, T&D losses are likely to be

7 The definition for JI is similarly worded, although it also includes reference to removals by sinks.


14

out of the control of the project participants, so reduction in T&D should not generate credits for mostsupply-side electricity projects (Kartha et. al. 20028).

The key words describing project boundaries in the Marrakech Accord, i.e. “under the control”“significant” and “reasonably attributable” are not defined, although they are all open to interpretation.Thus, the guidance provided by the Marrakech Accords is not specific enough to ensure consistent andcomparable decisions by project developers on determining what project boundaries for baselines andmonitoring JI/CDM projects should be.

Determining project boundaries is not necessarily a simple task (see e.g. OECD/IEA 2000). Greenhousegas mitigation projects can affect many different sources and gases. This impact can be direct or indirectand can occur on the project site or external to the project site. For example, a project’s fuel combustion,process emissions or sequestration are both on-site and direct. However, emissions associated with fueltransport to the project site, or electricity distribution from a project site are direct emissions, but occur off-site. Indirect on-site emissions could be caused by a change in operating characteristics of the project site(e.g. increased heating demand) or through preparation of the project site (e.g. flooding a site for hydrogeneration or deforestation on the project site). Indirect, off-site effects can also include the effect that aproject has on fuel demand and/or prices, or the GHG emissions generated by producing materials orequipment used on a project site.

The relative significance of sources of emission or mitigation can vary greatly by project type and duringthe life of the project. For example, the relative importance of different carbon pools can change during thegrowth of a forest. Establishing guidance on how to set project boundaries would therefore greatly improvethe comparability of different projects in the same sector.

2.3.1 Options to define “under the control” of the project participants

Project participants can control some, but not all, impacts of greenhouse gas mitigation projects. Forexample, they can control how much natural gas they use for power or heat production, but not the extentof leakage in the gas transmission and distribution network that supplies the project with gas. Similarly,distributors of energy efficiency appliances (e.g. energy efficient lightbulbs) can control how manyappliances they distribute, but not necessarily how much they are used. In general, project participantsshould be able to control direct on-site emissions. However, they may or may not be able to control off-siteor indirect emissions.

A standardised approach to defining “under the control” of project participants could be used. Thisdefinition could, for example, include either direct emissions only; or on-site emissions only; or all directemissions as well as emissions associated with the use of imported steam, heat or electricity andtransmission and distribution losses (for EE projects and distributed generation projects). Alternatively,“under the control” could be defined include any emission sources related to the project activity that can befinancially controlled by the project participants (WBCSD/WRI 2001).

Developing a standard approach to defining project boundaries will simplify work of project developers,and will therefore reduce transaction costs. It will also simplify the review of projects and verification ofemission reductions.

8 This study recommends including T&D losses in the project boundary for distributed electricity generation projects(and energy efficiency projects).


15

Recommendations:

� For most electricity/heat production projects, a definition of “under the control” thatencompasses direct on-site emissions, as well as emissions associated with any heat, steam orelectricity imported to the project site may be the most appropriate, as this includes aproject’s major sources, without requiring any confidential financial data9.� For energy efficiency and distributed generation projects, transmission and distributionlosses should also be included (Kartha et. al. 2002).

2.3.2 Options to define “significant” emission or enhancement sources

The relative importance of different sources of emission or enhancements in a JI/CDM project can varysubstantially between, but also within, project types. For example, the level of CH4 and N2O emissionsfrom fuel combustion will depend on the fuel source and technology type. For LULUCF projects, therelative importance of different carbon pools can vary substantially according to the climate zone in whichthe project is taking place (Figure 3). Therefore, a particular source/sink could be significant for oneafforestation project but not significant for another. This implies that prescribing what is a “significant”emissions/enhancement source cannot usefully be drawn up at the sector (or possibly even at a projectcategory) level.

Figure 3: Above and below-ground carbon stocks in different land types

0 100 200 300 400 500 600

Boreal forests

Tropical forests

Tropical savannas

Temperate grasslands

Wetlands

Deserts andsemideserts

Temperate forests

Croplands

Tundra

Carbon Stocks (Gt C)

VegetationSoils

Source: Data from IPCC SR LULUCF, 2000.

There are three main options to establish a general definition of a “significant” source of emissions (orenhancement). These are:

• Absolute emission (or enhancement) levels;

9 Some have suggested that a different definition should apply to JI projects, in order to exclude elements over whichthe host country has no control (such as the GHG-intensity of imported electricity or heat).


16

• Emission levels relative to the project’s (or baseline’s) total emission levels; or

• Emission levels relative to the largest single GWP-weighted GHG impact of a project.

Consistency and transparency between different projects in the same sector would be increased if theprojects used similar project boundaries. Work on establishing specific guidance on project boundarieswould thus be facilitated if the CDM EB could agree a general definition of, or formula to define, whatconstitutes a “significant” emissions source. Such a definition may, or may not, need to be differentiated bysector.

Alternatively, project developers could be responsible for identifying, quantifying and including allsignificant emission sources associated with their project in the project boundary. This would have theadvantage of avoiding the need to agree a definition of “significant” for particular project types. If projectdevelopers know from the outset (as they do for CDM projects) that an independent third-party will reviewthe emission benefits calculated from a particular project, it will not be in their interests to excludesignificant emission sources. Thus, this approach would not necessarily lead to large inconsistenciesbetween project boundaries chosen for different projects of the same type. (However, this approach maynot be as transparent as one applying an agreed definition of significant).

Recommendations:

� If an internationally-agreed definition of “significant” emission levels is needed, the mostappropriate approach may be to do so relative to the largest single estimated GWP-weightedGHG impact of a project, as it does not require either a subjective definition or quantificationof all emission sources, even if very minor. The World Bank uses this approach whencalculating the GHG impacts of their projects (WB 1998)10.� Alternatively, no concrete guidance of what constitutes “significant” could be drawn up.This would mean that developers of projects whose emission benefits will be verified by anindependent third party would be responsible for including all significant sources in theproject boundary.

2.3.3 Options to define “reasonably attributable” to the project

Individual greenhouse gas mitigation projects can have widespread impacts. However, not all impacts canbe directly attributed to a particular project and/or easily monitored. Defining what is “reasonablyattributable” to the JI/CDM project influences the extent of leakage or spillover from a project by decidingwhat is included or excluded from the project boundary. For example, it may be possible to attributeemissions caused by transport of a domestically-produced fuel to a project site to a particular JI/CDMproject. It is also possible to estimate these emissions. If these emissions are excluded from the projectboundary, the number of credits generated by the project may be overestimated. These emissions shouldtherefore be included in the project boundary if they are “significant”.

However, a project may have other effects on GHG emissions outside the geographical project boundary.For example, a greenhouse gas mitigation project involving a switch away from coal will reduce demandfor coal from the project site. It could therefore in theory reduce coal prices (although it is unlikely that onefuel switching project would significantly reduce the price of an internationally-traded commodity), and

10 The threshold used is to exclude any impact that is less than 10% of the largest individual GHG impact of theproject.


17

increase coal demand elsewhere. However, whether this effect is significant depends on the relative size ofthe project compared to world/regional/local coal demand, as well as a host of other factors such as the costand availability of coal and other fuels near the project site. Determining whether or not a project hasimpacts of this kind may be difficult, and quantifying such impacts would be even more so. Alternatively,it may be possible to quantify the GHG impacts of some indirect effects, such as the GHG intensity ofmaterials used to construct the project. For example, emissions from concrete used to build the shaft of awind turbine are “reasonably attributable” to the project using that wind turbine. However, counting suchemissions – if they are judged as “significant” – within the project boundary would increase the complexityof GHG reporting both at the project site and at the site where the project materials are produced as theseemissions may also be counted elsewhere.

Recommendation:

� For practical reasons, it may therefore be sensible to set limits around the definition ofwhat “reasonably attributable” is. This would include both a geographical component (e.g.relating to impacts on or adjacent to the project site) as well as a project activity component(e.g. relating to the project’s main activity).

2.4 Should guidance on project boundaries be the same for emission reductionand sink enhancement projects?

In LULUCF projects, the proportion of carbon in pools that are significant (see Figure 3) and attributableto the project, but difficult and/or costly to measure can be large. While estimating above-ground carboncan be done with reasonable accuracy and cost, monitoring the carbon sequestered in soils, roots andnecromass (dead biomass) is both costly and subject to considerable uncertainties (Hamburg 2000). Thus,including soil carbon, necromass - and to a lesser extent, roots - in project boundaries can significantlyincrease the cost of project monitoring. Of course, it will also increase the number of credits that a projectcan generate, although the relative importance of the different pools depends on the type of forest plantedand in which area.

Recommendation:

� Project developers should be given more choice in the definition of project boundaries forLULUCF CDM projects than for energy or industry projects. This choice could allow projectdevelopers the possibility of foregoing emissions credits, and associated costs of monitoringparticular parts of the project, for pools that they can demonstrate do not decrease as a resultof the project. This choice could be crucial in limiting the level of monitoring costs for small-scale LULUCF projects (such as those involving plantations for bioenergy).

2.5 How can the monitoring burden be reduced?

Small CDM projects are – by definition – less able to absorb transaction costs, including those associatedwith project monitoring. Reducing the monitoring burden for small CDM projects would reduce thetransaction costs associated with such projects and thus could help increase the number of such projectsthat are initiated.

The Marrakech Accords specifically include the possibility of developing simplified monitoringmethodologies for small-scale projects. There are a number of ways by which this could be done, (inincreasing order of simplification):


18

• Focussing monitoring efforts on larger emission sources or enhancements. This would imply thatsmaller projects could have a different definition of what a “significant” emission source (and thereforea different project boundary) to other project types.

• Using easy-to-monitor data, perhaps adjusted by using some sort of default values, rather than basingresults on data that are more difficult or costly to monitor. Thus it may be important to define proxiesthat can be used or easily developed for data that is unavailable or costly. For example, rather thanmonitoring how much electricity is actually distributed by a particular project, a standardised/defaultfactor could be applied to the project’s electricity generation to account for T&D losses.

• The use of published, default, emission/conversion factors (rather than measured factors) may also beappropriate, even for non-CO2 gases.

• It may also be worth exploring the use of sample populations, either for monitoring or verificationpurposes11. In such a situation, only a proportion of similar projects would be monitored and/orverified, and the GHG impact of the project estimated from the performance of the sample.

• The monitoring and/or verification frequency could be reduced for small projects. This could meanthat small CDM projects are monitored sporadically, rather than yearly12. For example, projects couldbe monitored in years 1, 2 and 7, with emission benefits for years 3-6 inclusive estimated from theextrapolated data. (It will probably be advisable avoiding extrapolating project results from data fromthe first year of project operation, particularly when the project involves the use/operation ofequipment of new technology types).

• Perhaps the simplest monitoring methodology may be to monitor only that the equipment installed inthe project continues to be in working order and is being used. This approach would only be applicableto the most environmentally friendly projects, such as zero-emitting technologies that are not incommon use in a particular region, Default factors would be used to estimate the output of suchprojects, rather than monitoring output directly. For small-scale renewable electricity projects based onintermittent energy resources, this may involve estimating a default load factor.

If many similar projects are undertaken in a region, it may be possible to group (or “bundle”) monitoringactivities together. This could reduce the costs of project monitoring substantially, although may increasethe error associated with calculating emission reductions from a group of projects. “Bundling” of projectscould be carried out for “small-scale” or other projects if the number of sample projects monitored shouldbe statistically significant. A multi-project verification exercise carried out for 31 comparable Swedish AIJprojects (DNV 2001a and b) estimates that multi-project verification activities could be 50-70% cheaperthan project-specific verification activities13. Case studies of greenhouse gas mitigation projects in Indiaalso indicated that bundling similar projects into groups 10 would “turn several types of small-scaleprojects into viable CDM projects” (Factorag 2001).

11 This approach would also usefully be applied to GHG mitigation projects that involve many diffuse sources: itwould be impractical to monitor every light bulb installed in an energy-efficiency project or every tree planted in aforestry project.12 However, the disadvantages of monitoring sporadically are that capacities for monitoring will not be built up, so thequality of monitoring may suffer.13 However, no substantiating information for these figures was presented in the document itself, as this informationwas considered to be confidential.


19

Recommendations:

� The implications of these different simplification strategies should be assessed, and moredetailed recommendations developed for each of the three types of small-scale projectseligible under the CDM.

� Projects should only be “bundled” together if they are technologically similar, located insimilar regions, have comparable baselines and monitor/report similar project indicators(DNV 2001a).


20

3. Other areas in which guidance may be useful

International guidance provided on other issues could also be useful in increasing comparability andtransparency between projects, and reducing transaction costs. These include tasks such as:

• What a monitoring plan should contain and how data should be recorded; and

• The relationship between the frequency of monitoring, reporting and crediting.

In addition, guidance may also be usefully prepared on:

• Simplifying or streamlining monitoring procedures for small-scale CDM projects.

3.1 What a monitoring plan should contain

The Marrakech Accords calls for a monitoring plan that collects and archives all data relevant todetermining the project’s baseline, monitoring its performance and determining the GHG benefits of theproject (see Annex A). Guidance on developing this plan and on monitoring the performance of projectswill need to be a practical "how-to" document designed for use by project developers and participants.Monitoring guidance can range from very broad guidance, such as what constitutes a “significant”emissions source to detailed technical recommendations, such as where to locate meters or how to readthem.

The Marrakech Accords include guidance on what the monitoring plan for JI/CDM should include (seeAnnex A). Each of these points, such as “quality assurance and control procedures” could usefully beelaborated on in a monitoring plan or any specific monitoring guidance to be developed. Moreover, otherinformation not mentioned in the Marrakech Accord outline of what a monitoring plan should include mayalso be needed in such guidance, e.g.:

• Information on pre-project (baseline) situation. This information will be needed by the OperationalEntities in order to calculate the credits generated by a project.

• An outline of the organisational and management requirements for monitoring, such as in whichlanguage data should be recorded, archiving of monitoring data, who is responsible for monitoringproject performance, and what (if any) training will be provided to ensure that this monitoring iscarried out.

• An indication of monitoring frequency and reporting.

• Explanation of how emission reductions/enhancements due to the project were calculated by using thedata monitored with the monitoring methods described (e.g. instructions on how to “translate”monitored project data to GHG emissions)

• What information is likely to be needed for verifying emission credits, for example fuel bills,documentation of system losses, indication of project context etc.

In addition, information on the sustainable development aspects of the project may need to be monitored inorder to facilitate host country judgements on whether or not a project proposed or underway does


21

contribute to their sustainable development. Monitoring plans written for PCF projects include specificindicators that projects need to meet to show that they are contributing to sustainable development.

3.2 Frequency of monitoring, reporting, and verifying information

Project developers will need guidance on how often to monitor, record, aggregate (e.g. averages across agiven period) and report data. There may or may not be a significant difference between the frequency withwhich a project needs to be monitored, and the frequency with which monitored data needs to be recordedand reported. For example, an electricity project using continuous emissions monitors or monitoring hourlyoutput will not need to report all data collected, but project participants will need guidance on how often toreport data collected on a continuous or very frequent basis.

3.2.1 Monitoring and reporting frequency

The frequency of monitoring should be determined by answers to the following questions:

• What frequency of monitoring will give an accurate picture of project performance?

• With what frequency are credits from the project needed?

The frequency of monitoring that will give an accurate picture of project performance will vary by, andwithin, project type. For example, although re/afforestation projects increase below-ground carbon as wellas above-ground carbon, increases in below-ground carbon can be both slow and gradual (depending onthe species). Periodic (e.g. every 5y) monitoring of below-ground carbon will therefore give an accurateestimation of its accumulation in the years when it is monitored, as well as in the years betweenmonitoring. It will therefore be most cost-efficient, while also remaining environmentally prudent, tomonitor such information on a less than annual basis.

However, parameters in other project types that are inherently more variable will need to be monitoredmore frequently. For example, a project’s level of output may need to be monitored more frequently thanthe characteristics of its raw materials or input fuel. An appropriate monitoring frequency may even varywithin a project category. Energy efficiency projects that reduce a constant load by a constant amount mayonly needing limited (sporadic) monitoring, whereas a project reducing a variable demand by a variableamount would need more regular monitoring in order to determine accurately the energy saved by thatparticular project (IPMVP 2000).

3.2.2 Verification frequency

According to the Marrakech Accord, it is the operational entity that determines the emission reductionsfrom CDM projects. The Accords also stipulate that operational entities (OE) need to submit an annualactivity report to the Executive Board. Furthermore, credits can only be issued after the OE has submittedits verification and certification reports. These reports cover emission reductions “during the specified timeperiod”. However, the Accords do not indicate whether all projects need to be verified or certifiedannually.

It may therefore be possible for project participants of at least some project types to choose a less-than-annual (e.g. once every 2 or 3 years) verification schedule. This would mean that although the projectwould continue generating emission credits on a regular basis, credits may only be issued ever few years.


22

A less-than-annual verification process would not necessarily need to alter the monitoring and reportingschedule used within the project itself, although it would mean that project operators responsible formonitoring data would need to keep data over a longer period of time.

3.2.3 Frequency/cost trade-offs

There will be a trade-off between the time/cost of frequent monitoring and reporting, and the time/costneeded to verify a project’s emissions/enhancements (Table 8). Verifiers will also need to have access tomonitoring data, and may also wish or need to undertake limited monitoring themselves as part of anauditing procedure. Verification is an important exercise to ensure the environmental credibility of thecredits generated. It can also highlight significant differences between estimated and audited emissionreductions. For example, DNV (2001b) indicate that the reported mitigation effect of 31 projects was 746ktCO2, but the audited (verified) mitigation effect for the same projects was 13% lower at 647 kt CO2.

Less-than-annual verification should lower the fixed costs of verification (e.g. travel to project site),although it may increase the variable costs (e.g. of identifying data anomalies). A less frequent verificationschedule would mean that project developers would receive emission credits (and associated revenue) inbunches, rather than yearly. Moreover, a delay between a project generating credits and a project developerreceiving credits would normally be expected, because of the time needed for verification, certification andissuance of credits. Thus, the cost/ease of verification may also need to be considered when establishingguidance for the frequency of monitoring and reporting data, although such considerations should also takeinto account the cost differences in more frequent monitoring and reporting.

3.3 Areas for detailed/sectoral monitoring guidance

Specific monitoring guidance by project type/category, in addition to the ground rules for monitoringguidance described above, will facilitate the work of both the project developer and the project operator indetermining what to monitor, how and when.

Specific guidance is likely to be needed on the key technical and organisational aspects of projectmonitoring, namely:

• Which sources/gases should be included in the project boundary (either a list, or a method by which todetermine the project boundary);

• A detailed list of parameters that need monitoring, e.g. kWh distributed;

• Which monitoring method(s) and/or equipment should be used;

• Other reporting requirements, e.g. indicating the status of different data points (measured, monitored,derived, default value etc.);

• The frequency with which parameters relevant for the estimation of the project’s emissions need to bemonitored. This is likely to vary considerably for different parameters, and may also vary dependingon different project types.

• When the project needs to be monitored. It is important for LULUCF projects that monitoring occursat the same time of year. Consistency in timing of monitoring may also be important for projectsestimating benefits on the basis of sample populations.


23

• An outline of the organisational requirements for monitoring, such as how, where and in whichlanguage data should be recorded. This is important because data may be recorded over a long timeperiod, and may need to be accessed throughout the life of the project. It is also important that the databe recorded in such a form as to be easily accessible and understandable by a verifier. The allocation ofmonitoring responsibilities between the different project participants is also needed.

• Recommendations on reducing errors and uncertainties (including recommendations on how togenerate a sample population and extrapolate emission reductions/sink enhancements from thispopulation, the use of trained personnel to monitor different items) and recommendations on how todeal with errors/uncertainties (such as tradeoff with number of credits).

• How long a project should be monitored for. Depending on how crediting for LULUCF projects is setup, guidance may be needed on whether a LULUCF project needs monitoring past its creditinglifetime.

• How leakage, spillover and transboundary effects will be determined.

• Units in which to monitor and report. Specifying frequently commonly used units for reporting (suchas volume units for natural gas) may reduce reporting errors (WRI/WBCSD 2001). Specifyingcommon reporting units will facilitate verification.

In addition, it may be necessary to provide the project operators with some guidance or training in how tooperate the monitoring and reporting system set up for the project (DNV 2001b). For example, informationrequired to calculate emission reductions will need to be monitored on a calendar year basis, whereas otherinformation the project operators may need to monitored or report for economic reasons may be on afinancial year basis.

Since this guidance is technical, it may be most appropriate for it to be drawn up by experts, e.g. thoseserving on an advisory panel to the CDM EB. A useful starting point may be guidance for JI/CDM projectsdrawn up by the Japanese (WBG 2001) and Dutch governments (ERUPT and CERUPT, EZ 2001) and thework on monitoring energy efficiency and forestry projects drawn up by LBL (Vine and Sathaye 1999,Vine et al. 1999).

Other guidance can also provide useful input to development of detailed monitoring methods and guidance.For example, IPCC guidance (IPCC 1997 and 2000) outlines methodologies for determining emissionsfrom different sources; develops decision trees to determine which the most appropriate method is; andlists detailed technical data (such as emission factors, conversion factors) that may be useful whencalculating the emissions associated with a project. It also provides management/organisational guidancefor the development of quality assurance/quality control (QA/QC) procedures. The IPMVP (IPMVP 2000)provides an overview of best practice techniques in determining the results of energy efficiency projects.National guidelines e.g. to determine the mitigation effects of voluntary actions by companies, have alsobeen developed and may be relevant. In addition, much experience has been gained using themethodologies developed by Winrock (MacDicken 1997) to calculate changes in carbon stocks fromforestry projects. ISO 14000 and 9000 may also provide useful input, as they make recommendations onthe responsibilities and qualifications of the different team-members used in verification teams and onQA/QC procedures.


24

3.3.1 Specific guidance recommendations

Because the relative importance of different sources of emissions or enhancements in JI/CDM projectsmay vary by project type, it may be sensible to use decision trees in order to guide decisions on drawing upproject boundaries. This would reduce the number of separate monitoring guidance documents that wouldneed to be set up. An example of a decision tree for deciding the boundary of an electricity project isoutlined in Figure 4.

Figure 4: Example of a decision tree determination of project boundaries (electricity orheat generation project)

Does project use biomass or fossil fuelon site (even as backup fuel?)

No

Yes

Emissions = 0. No need to quantify.

Quantify emissions from fuel combustion.

Are CH4 and N2O emissions from eachfuel > [x], >[y%] of project emissions?

No

Yes

Exclude from project boundary.

Include in project boundary.

Does the project involve on-site storageof oil or gas (or derivatives)?

No

Include in project boundary if preliminarycalculations estimate that emissions/y> [x], [y%]. If not, include in leakage estimate.

Yes

Emissions = 0. No need to quantify.

Does the project increase road transportof fuels to project site?

No

Yes

Exclude from project boundary.

Is fuel source transported > [X km][Y km/y] to project site?

NoExclude from project boundary.

YesInclude in project boundary if preliminary calculations estimatethat emissions/y > [x], [y%]. If not, include in leakage estimate.

1.

2.

3.

4.

Does project use > [x kWh] [y%] ofgeneration for own use?

No

Yes

(does not influence boundary)

Does project export electricity to grid? No

Yes

5.

6.

Include own use in project boundary.

Use unadjusted output figures.

Do preliminary estimates indicate thatthe project will change T&D lossesor that T&D losses will be > [y%]?

No

Exclude from project boundary,but include in leakage estimate.

Yes

Include in project boundary.


25

3.3.2 Deciding which aspects of a project to monitor

Not all aspects of greenhouse gas mitigation projects will be monitored on a continuous basis. Therefore,project performance will need to be extrapolated from the use of indicators of project performance.

A definition of project boundaries (whether it appears in guidance on project monitoring or on baselinedevelopment) will help project developers assess what information does not need to be monitored.However, it may not be precise enough to define what exactly should be monitored. Thus, monitoringguidance will also need to recommend what indicators of project performance should be, and how theyshould be monitored.

As shown by Table 2 to 5, what needs to be monitored is determined by the project boundary and varieswidely by project category. However, defining the project boundaries does not in itself completely definewhich indicators need to be monitored in order to measure project performance. For example, there is noagreement on what indicator should be used to define project output for an electricity/heat project. Even foran electricity generation project, for which monitoring should be relatively simple, there can be significantdifferences between electricity generated by a project, energy sent out to the grid (i.e. excluding own-use)and electricity received by the user (i.e. taking into account losses from transmission and distribution).Consistency, comparability and transparency will be increased if guidance is given on what “type” ofelectricity (generated, sent out, used) should be monitored. (This should be straightforward. For example,the baseline recommendations for GHG mitigation projects in the electric power sector in Kartha et. al.2002 are based on generated electricity. Technical T&D losses are only taken into account for energyefficiency projects and distributed generation projects).

The use of decision trees may also be useful in determining the activities and parameters that need to bemonitored to assess project performance, as well as outlining how (in which units) they should bemonitored. This information is needed before the project is set up in order to allow cost-effectivemonitoring systems to be designed or located in the most appropriate locations. Defining the units in whichproject output should be monitored may be less obvious for projects in other sectors (e.g. industry), as thebaseline could be expressed in terms of emissions/tons of intermediate product produced (rather thanemissions/ton of final product) (OECD/IEA 2000).

Once it has been decided what aspects of a project need to be monitored, the next step is to assess howsuch monitoring should occur. In some cases, such as forestry or energy efficiency projects, more than onemethod could appropriately be used to monitor individual parameters of project performance. Projectdevelopers could choose among methods, and their choice might depend on the type of project, therequired accuracy of estimate, as well as the funds available for monitoring.

Once the indicators of project performance have been monitored, they then need to be "translated" into aGHG emissions equivalent by using an emissions factor. This will be a more or less simple task dependingon the type of project. For example, an electricity/heat production project will need to calculate the GHGproduced by the input fuel to generate electricity. This is a trivial task for projects using zero-emitting fuels(if no other fossil fuel sources are included in the project boundary), and should be able to be donerelatively easily for other project types by using the emission calculation methodologies developed by theIPCC. “Translating” carbon monitored in forestry projects to CO2 is also a trivial process.


26

4. Lessons learned from experience in monitoring/verifying projectperformance

Some of the organisational lessons learned from monitoring greenhouse gas mitigation projects havealready been incorporated into the CDM and JI project cycles as laid out in the Marrakech Accord. Forexample, the requirement that a monitoring plan needs to be set up at the project design stage shouldenable monitoring considerations to be included at an early stage in project planning. This should thereforefacilitate the gathering of relevant data at little or no extra costs by e.g. ensuring that meters are placed atappropriate locations. Ensuring that emission reductions are verified by an independent third party shouldalso increase confidence in the validity of issued emission reductions. Other lessons learned frommonitoring greenhouse gas mitigation projects already underway, such as the importance of trainingproject operators in monitoring and reporting climate-related impacts of the project (as distinct from theeconomic and/or physical output of the project) should also be incorporated.

There is already some experience with developing and/or applying monitoring plans for greenhouse gasmitigation projects. Some monitoring guidance has been developed at the sector and/or project categorylevel specifically to monitor the performance of JI/CDM (or other climate mitigation) projects. Thisguidance has generally been drawn up by individual governments (e.g. EZ 2001, WGB 2001) ororganisations (e.g. Vine and Sathaye 1999, Vine et. al. 1999, MacDicken 1997, Martinot 1998, DNV1999a). The IPCC has also drawn up some very generic guidance for LULUCF projects (IPCC 2000).

Monitoring plans have also been established for a few AIJ projects (e.g. Ravindranath and Bhat 1997,Brown et. al. 1999), and the monitoring/verification process analysed for others (e.g. DNV 1999b, DNV2001). However, many AIJ projects were not systematically monitored (e.g. DNV 2000), so experiencewith monitoring GHG mitigation projects is not as large as could be expected. A handful of PrototypeCarbon Fund (PCF) projects, all of which have a monitoring plan, have also been developed (e.g. PCF2001a and 2001b).

Existing monitoring plans and guidance documents can differ substantially on key issues, such as how todefine project boundaries. This section summarises experience with and/or planning for project monitoringin the electricity, energy efficiency and forestry sectors, including on project boundaries, monitoringmethods and cost. It then highlights differences between different available guidance documents, and raisesquestions and develops recommendations on monitoring guidance for JI/CDM projects in the electricityand forestry sectors.

4.1 Project boundaries

General guidance on project boundaries for GHG mitigation projects has been developed by a number ofdifferent organisations. These were all developed before the Marrakech Accords were agreed, and so ofcourse could not take its definition of project boundary into account. While the guidance in the MarrakechAccords will supercede guidance previously developed for JI/CDM projects, it is interesting to note thevariation found in different guidance documents (Table 1).


27

Table 1: General recommendations on which impacts to include in the project boundary

Guidance Sectors covered Definition of boundaryMarrakechAccords

All JI projects and emissionreduction CDM projects

Boundary shall include all sources “under the control ofthe project participants that are significant and reasonablyattributable” to the project.

OECD/IEA(2000)

Electricity generation,energy efficiency(lighting), cement, iron &steel, transport

Guidance varies by sector and project type, but generallysuggests including major (but not all) direct on-siteemissions, as well as emissions associated with electricityused in the project.

ERUPT(2001)

Fuel switch (in electricity,heat or CHP projects),CHP, landfill gas recovery,forestry

Direct on-site and off-site emissions that can becontrolled or influenced by the project. In general,emissions related to activities one step downstream orupstream are also included.

WRI/WBCSD(2001)

General Direct GHG emissions from sources owned or controlledby the reporting company. Emissions from the generationof imported/purchased electricity, heat and steam.

WBG(2001)

Energy efficiencyimprovement in steelworks,refineries, fossil fuel powerplant; natural gas CHP;reforestation

Include direct emissions (removals) from the project’smain activity, as well as those not related to the project’smain activity but that are not “small enough to beignored”. Include some indirect project impacts in projectboundary.

WorldBank(1998)

Several project types in theenergy, transport, industry,and LULUCF sectors.

Guidance suggests excluding any impact that is less than10% of the largest individual GHG impact of the project.

This difference in approach to defining project boundaries between different guidance documents can beillustrated by examining detailed recommendations on what to include and exclude in the project boundaryfor certain project types.

Table 2 outlines different recommendations for an electricity, CHP or heat supply-side project. The onlyaspect on which there is positive consensus is that CO2 from the fossil fuels used to generateelectricity/heat should be included in the project boundary. There also seems to be implicit agreement (assome guidance documents do not refer to these as possible sources) that emissions from site preparationand carbon embodied in construction materials should be excluded from the project boundary.

However, there are mixed opinions as to whether or not indirect emissions or non-CO2 gases should beincluded in the project boundary. Whether or not these “other” sources are included may only make a fewpercentage points difference in the number of credits generated by a project, but could make monitoringsignificantly more expensive. These include CH4 and N2O from fossil fuel combustion, which togetherrepresent less than 1% of direct GHG emissions from the combustion of fossil fuels to generate electricity(OECD/IEA 2000). The different guidance documents are also not agreed on whether or not to includeemissions associated with the transportation of fuels to the project site and of fuel storage on-site (althoughboth sources are "reasonably attributable" to the project activity). In some cases, emission from transport ofbiomass or storage of natural gas could be significant, although data constraints may make them difficult tocalculate.


28

Table 2: Comparing project boundary recommendations (electricity/heat supply)

Source Included in boundary Excluded from boundary Project type

ERUPT (andCERUPT draft)(EZ 2001)

On-site combustion of fossilfuels, own use of electricityand heat

Combustion of fuels in back-upboiler

CHP

WBG (2001) On-site combustion of fossilfuels, emissions fromtransportation and storage offuels on project site

Mining and processing of fuelsand construction materials.Operation of constructionmachinery, construction ofinfrastructure for heat supply,change of biomass caused bylandcover change.

CHP (gas)

OECD/IEA(2000)

CO2 and CH4 emissions fromon-site combustion of fuel forelectricity generation

On-site emissions of N2O, all off-site and indirect effects

Grid-connectedelec. gen.

WB (1998) CO2 emissions from on-sitecombustion of fuel forelectricity generation.

On-site emissions of CH4 andN2O. Losses from transmissionand distribution.

Grid-connectedCHP.

IGPO (2000) CO2, CH4 and N2O emissionsfrom on-site combustion offuel for electricity generation,own use of electricity/heat

All off-site and indirect effects Electricity orheat projects

Table 3 indicates the project boundaries actually used in selected electricity/heat greenhouse gas mitigationprojects (e.g. AIJ or PCF projects). These boundaries, unsurprisingly, also vary. PCF projects have verystrict monitoring and reporting requirements in general. This is illustrated by the fact that one projectexcludes increased electricity use due to "development" (i.e. increase in electricity use over business-as-usual due to readily available electricity) from its project boundary. Another PCF project uses hourlydispatch data to assess the GHG-intensity of electricity displaced by the project. While such detailedmethods increase the accuracy of estimating the climate mitigation effect of the project, they also increasethe monitoring burden, both before and during the project. However, even different AIJ projects havedifferent project boundaries. Although many AIJ projects in the electricity/heat sectors exclude CH4 and/orN2O, these sources are included for some projects.


29

Table 3: Project boundary assumptions in selected climate mitigation projects(electricity/heat)

Source Included in boundary Excluded from boundary Project type

PCF (2001b) Transport and combustion ofpetroleum fuels forelectricity generation andlighting purposes to projectarea

Increased electricity use dueto “development”. Non-CO2

gases produced bycombustion and transport.

Small elec genproject (hydroand diesel)

PCF (2001a) Hourly net generation (localgrid collects figures onhourly generation by source)

Run of riverhydro (elec.gen.)

Jordan/E7 AIJproject (UNFCCC2000)

CO2 emissions from oil usedfor power generation,including own use.

CH4 and N2O emissionsfrom fuel combustion, alloff-site emissions

Fuel efficiencyimprovement

Poland/NetherlandsAIJ project atByczyna

CO2, CH4 and N2Oemissions from boiler fuelcombustion

(All other emissions) Fuel efficiencyand fuel switchin heatingsystem.

Martinsen et. al.(2001)

Physical boundary of theheat production facility.

Emissions from fuelproduction andtransportation to project site.

Heat production(biofuels)

Table 4 outlines recommendations from different guidance documents on what to include in projectboundaries for a reforestation project. There is greater consensus than for electricity/heat projects, but notcomplete agreement, on what to include in a project boundary. For a reforestation project, all guidancesuggests including the carbon contained in above-ground biomass, and excluding emissions from e.g.transporting seedlings and wood products. However, there is disagreement on whether or not to monitorsome significant carbon pools, such as soils and wood products, although whether soil carbon provides a"significant" contribution to GHG sequestration from a project depends on the soil type, previous land usesand the type of forest that is planted (Figure 3).


30

Table 4: Comparing project boundary recommendations (reforestation project)

Source Included in boundary Excluded from boundary

WBG (2001) Above and below-ground biomass, litteraccumulation, emissions from soil carbon,preparation of land for planting, fertiliser use,reversal of sequestration from fires,deforestation etc.

Fossil fuel combustion for raisingand transporting seedlings,operating machinery, transportingwood products, land use on landadjacent to project.

ERUPT Sink enhancement pools (which pools are notspecified), fertiliser use, fossil-fuel-relatedemissions related to land preparation, nursery,planting,

(instructions unclear as to whetherdirect and indirect off-siteemissions should be included)

Vine et. al.(1999)

All carbon pools, possibly C content ofharvested wood products, “and can be expandedif warranted”, temporal boundaries

(Emissions related to landpreparation and planting, fertiliseruse not mentioned)

IPCCSRLULUCF*(2001)

Trees and wood products should be quantifiedand monitored. It is recommended that roots andsoil carbon are also quantified and monitored.Dead biomass may need to be measureddepending on project type.

Herbaceous biomass (vegetation)excluded from boundary.(Emissions related to landpreparation and planting, fertiliseruse not mentioned)

Hamburg(2000)

Living above-ground biomass, below-groundbiomass (roots), soil carbon (if stock declines).

Soil carbon (if increased byproject), necromass.

Tipper et. al.2001

All carbon pools that increase for which creditsare claimed, all carbon pools that decrease. Mayneed to include non-CO2 gases. May need toinclude fossil fuel emissions from e.g. irrigation.

Carbon pools that increase but forwhich credit not claimed.

*The suggested boundary here is for a re/afforestation project involving plantations (but not short-rotation energyplantations). The recommended boundary for other reforestation projects e.g. agroforestry projects is different.

Table 5 outlines different project boundaries actually used for different LULUCF projects. As for theelectricity/heat example, there is agreement from all different guidance documents that the one majorsource of GHG mitigation from forestry projects (carbon contained in trees) should be included in theproject boundary. The guidance documents that mention fertiliser use suggest including it in the projectboundary (but most guidance documents do not mention fertiliser). However, there are differentrecommendations for almost every other aspect, including significant carbon pools such as roots, soilcarbon, dead biomass or harvested wood products. How to deal with the possible reversal of carbonsequestration by fires, pests, etc. (the "permanence" issue) is also not included in all monitoring guidancedocuments. Few guidance documents indicate whether or not emissions associated with forestry projects(such fossil fuel emissions associated with land preparation and planting) should be included. There is alsoa wide difference in what is included in the project boundaries for individual (mainly AIJ) forestryprojects.


31

Table 5: Project boundary assumptions in selected climate mitigation projects(reforestation)

Source Included in boundary Excluded from boundary

Ravindranath &Bhat (1997)

Standing tree biomass, litter, soil carbon,annual C uptake, wood extraction, woodend uses, soil and litter decompositionrates, root biomass accumulation

(Emissions related to land preparationand planting, fertiliser use notmentioned)

Brown et. al.1999

Above-ground woody biomass (trees),below-ground biomass (roots), deadbiomass (standing and lying) vegetation,litter

Soil carbon. (Emissions related toland preparation and planting,fertiliser use not mentioned)

Klinki AIJ(UNFCCC 1998)

Above and below-ground biomass Soil carbon, emissions fromharvesting.

Scolel Té project(IPCC 2001)

Trees, soil All other carbon pools. (Emissionsrelated to land preparation andplanting, fertiliser use not mentioned)

4.2 Frequency of monitoring

There is only limited guidance available on how frequently to monitor different aspects of a project’sperformance. As for the different guidance on project boundaries, guidance on monitoring frequency alsodiverges.

Table 6 outlines recommendations for the monitoring frequency of two forestry projects. Therecommended monitoring frequency is quite different for important carbon pools, such as above-groundwoody biomass and below-ground biomass (roots). Provided that the project site was not subject todisturbances, the monitoring frequency may not affect the total number of credits generated by a project,and may therefore be judged as unproblematic. However, the timing and frequency of monitoring mayaffect the years in which credits are allocated. For example, if biomass growth (and carbon accumulation)is extrapolated linearly between starting, mid and end-points, the growth of carbon stocks will beoverestimated at the beginning of the project and underestimated towards the end. This slight front-loadingof credits may make be the project more economically attractive (Ellis 2001).


32

Table 6: Selected recommendations on frequency of monitoring activities for differentcarbon pools in LULUCF projects

Suggested timing (y)Parameter to bemonitored “Typical forestry mitigation

project”*Guaraqueçaba Climate ActionProject, Brazil*

Soil carbon Y0, every 2/3 years, end of project Years 1, 3, 5, 10, 15, 20, 30, 40Above-ground woodybiomass

Y0, mid rotation, end rotation Years 1, 3, 5, 10, 15, 20, 30, 40Remote sensing will be used in years25 and 35

Standing dead biomass -- Years 1, 3, 5, 10, 15, 20, 30, 40Litter/slash Y0, every 5 years Years 1, 3, 5, 10, 15, 20, 30, 40Annual C uptake Annually or periodically --Extraction of wood Annually, end of rotation (not applicable to this project)End-uses of wood Annually, end of rotation --Soil/litterdecomposition rates

Annually Every 5 years

Root biomass Y0, end of rotation Every 5 yearsRates of land-usechange, socio-Economic indicators

-- Y0, every 5 years

Source: Ravindranath & Bhat, 1997 (for a “typical forestry mitigation project”), Brown et. al. 1999, Brown &Delaney 1999 (for Guaraqueçaba Climate Action Project, Brazil)

Table 7 outlines the recommended monitoring frequency for a fuel-switch AIJ project. This tablehighlights the different frequency with which different parameters in one particular project should bemonitored. While some aspects of the project (i.e. generation of electricity and heat) are monitoredcontinuously, other aspects are monitored daily or yearly. Moreover, double copies of the project outputare kept: in electronic form (continuously) and paper form (weekly). Recommendations on monitoringfrequency from other electricity supply projects are different. For example, one PCF project (PCF 2001b)requires monthly output reports. Another (PCF 2001a) indicates that hourly output needs to be recorded.The mitigation effect of AIJ projects, when reported to the UNFCCC, are done so on an annual level (seee.g. UNFCCC 2001b), although these reports do not indicate the frequency of reporting compared to thefrequency of monitoring.


33

Table 7: Monitoring frequency and responsibilities for Quercus AIJ project

Data Recordingfrequency

Format(paper/

electronic)

Where kept Responsible Datarecordsstored

Generated electricity Continuous Electronic Boiler house Boiler housemanager

2 years

Generated electricity Weekly Paper Head office Boiler houseoperator

2 years

Generated heat Continuous Electronic Boiler house Boiler housemanager

2 years

Generated heat Weekly Paper Head office Boiler houseoperator

2 years

Bio fuel consumption Daily Paper Boiler house Boiler houseoperator

2 years

NCV of bio fuel Daily Paper Boiler house Boiler housemanager

2 years

Gas consumption Daily Paper Boiler house Boiler houseoperator

2 years

NCV of gas Annually Paper Boiler house Boiler housemanager

2 years

Source: Martinsen et al. 2002

Monitoring guidance for LULUCF projects may also need to specify the length of time a project should bemonitored for. This is because the guidelines for CDM projects agreed at COP7 do not cover LULUCFprojects and because forestry projects sequester carbon over a long period of time. Moreover, since carbonsequestration may be reversed, projects may need to be monitored for a longer period than their creditinglife to ensure that the credits generated by forestry projects represent real and long-term climate benefits.


34

5. Conclusions

Monitoring activities will need to be carried out at different times and for different purposes during thepreparation and implementation of a greenhouse gas mitigation project. For example, pre-projectmonitoring may be needed to define a project's baseline and monitoring during the project will be neededto quantify a project’s greenhouse gas impacts (and therefore the number of credits it can generate).Monitoring may also be needed for other project-related functions such as assessing the importance ofleakage or whether or not the crediting period for a project should be extended. This paper focuses onmonitoring the performance of a greenhouse gas mitigation project and on defining the boundaries of whatneeds to be monitored.

The CDM's Executive Board has been tasked with setting up detailed guidance on monitoring, baselines,project boundaries, as well as other issues relevant for CDM projects. Readily available monitoringguidance will facilitate the task of JI/CDM project developers by giving clear instructions, and cantherefore reduce the cost to project developers of establishing a monitoring plan (this is also true fordomestic GHG mitigation projects). Clear monitoring guidance can also facilitate verification of emissionreductions.

The Marrakech Accords indicate that the monitoring plan for a JI/CDM project has to be set up as part ofthe project design document. Thus, monitoring-related costs will occur up-front (even before a project hasbeen registered) as well as during a project’s operation. The costs associated with developing a monitoringor monitoring/verification plan can be significant, e.g. $20,000 for a “first-of-a-kind” project. The on-goingcosts of monitoring project performance can also be significant, but vary by project type and design.

Developing a project’s baseline and monitoring requirements are closely linked. For example, allparameters included in a project’s baseline will also need to be included in the definition of a projectboundary, and monitored during the project’s crediting period. Therefore, developing guidelines for projectmonitoring and baselines may need to be carried out in parallel in order to ensure consistency. Forexample, if a particular emissions source is excluded from a project's baseline, monitoring reductions inemissions from this source should not generate emission credits. Developing monitoring guidelines willalso influence (or be influenced by) any guidelines on how to report and/or verify the impacts ofgreenhouse gas mitigation projects.

Monitoring guidance can range from very broad guidance, such as what constitutes a “significant”emissions source to detailed technical recommendations, such as where to locate meters or how to readthem. Thus, project developers would benefit both from “ground rules” as well as specific methodological,management or organisational guidance.

Some key questions need answering before the ground rules for any detailed (e.g. sectoral or project-category) monitoring guidance can be established. The answers to these questions are inter-linked andbegin to establish the ground rules for project monitoring guidance. These questions need to be addressedrapidly in order to allow monitoring plans, and therefore JI/CDM projects, to move ahead. The answerswill influence the time and cost associated with developing, applying and verifying monitoring guidance.These are:

• Whether monitoring guidance should allow for a choice of monitoring methods. Given that differentmethods can be used to calculate emission baselines, that particular emissions can be ably monitoredby different methods, and that some flexibility to use less complex monitoring methods could reducemonitoring costs, this paper recommends that a choice of monitoring methodologies should beallowed.


35

• Whether individual project developers should have some flexibility in deciding the project boundary.How the project boundary is defined affects which sources of emissions and enhancements areaccounted for, and therefore the importance (or not) of leakage. The project boundary therefore alsoaffects which emission sources/sinks are included in a project’s baseline. In order to promoteconsistency and comparability between different projects of a similar nature, it may be appropriate torequire similar projects to have comparable project boundaries. This paper recommends encouragingconsistent project boundaries either by further clarifying the definition of project boundary outlined inthe Marrakech Accord, or by devolving the responsibility to define what is “significant” “reasonablyattributable”, and “under the control” to the project developers and verifiers14.

• Whether any general cost/accuracy tradeoffs should be recommended. This paper recommends that“incremental” monitoring costs for larger GHG mitigation or sequestration projects be keptmanageable by ensuring that a very high level of precision/confidence is not required for projectmonitoring data if there are significant uncertainties in a project’s baseline level or data. It alsorecommends that simpler monitoring methods should be allowed in return for a conservativeestimation of credits generated from the project.

• How monitoring procedures for small-scale projects could be simplified. Simplifications could includereducing the frequency of monitoring/verification; and calculating project benefits by using easy-to-monitor data, default emission factors and narrow project boundaries. For the most environmentally-friendly projects, such as renewable-based electricity generation, highly simplified monitoring couldbe limited to ensuring that the equipment installed continues to operate.

Once the “ground rules” have been set, more specific guidance will be needed by project category or type.This type of guidance includes:

• What a monitoring plan should contain. The Marrakech Accords outline some parameters that areneeded in a monitoring plan, but a complete list needs to be drawn up.

• How frequently projects should be monitored and the relationship between the frequency ofmonitoring and the frequency of reporting and verification. This is important because the frequency ofmonitoring, reporting and verification influence the transaction costs of the project.

A project’s monitoring plan may need to contain items in addition to those identified in the MarrakechAccord. These include “technical” data such as a detailed list of all sources of emissions and enhancementsincluded in the project boundary; monitoring methods chosen for different sources and gases; methods onhow to assess or address errors and uncertainties; information on the pre-project situation, and anindication of monitoring frequency and reporting. In addition, the monitoring plan will also need to includean outline of organisational/management requirements for monitoring (e.g. where and how data should bestored); who is responsible for monitoring what; and what information is likely to be needed for verifyingcredits.

Some experience with developing monitoring plans and monitoring greenhouse gas mitigation projectsexists. For example, some emission reduction and sink enhancement AIJ projects have been monitored,although many AIJ projects have not been monitored regularly. In addition, detailed and extensivemonitoring and verification plans have been drawn up for each Prototype Carbon Fund project. Variousgovernments and organisations have also drawn up guidelines of varying levels of detail on how to monitor

14 It will not be in the interests of the project developers to omit significant emission sources from the projectboundary if they know that independent verifiers will subsequently examine their boundary choices and can withholdor delay credit issuance if they are in disagreement.


36

greenhouse gas mitigation projects. This experience, and the lessons learned from it, will be useful input toany detailed, methodological monitoring guidance that is drawn up for CDM and JI projects.

Moreover, detailed guidance relevant to monitoring the benefits of greenhouse gas mitigation projects hasalso been drawn up for other purposes. This includes the IPCC guidance on good practice in inventorycalculations, the International Performance Measurement and Verification Protocol for energy efficiencyprojects and the WRI/WBCSD Greenhouse Gas Protocol, developed for corporate accounting. In additionto these international standards, national or regional standards for measuring and monitoring have alsobeen developed for some project types. Using such standards to measure and monitor the impact ofJI/CDM projects should be encouraged, as it should increase transparency of monitoring, and decrease thecosts of developing monitoring plans.

Comparisons of the detailed guidance on project monitoring drawn up by different bodies, shows thatdifferent organisations give different recommendations on what to monitor, and when. (Not all guidancerecommends methods of how to monitor particular parameters). Differences in project boundaryrecommendations reflect the implicit decisions by different guidance developers of what is or is not asignificant emissions/enhancement source that can be monitored cost-effectively. For example, forelectricity/heat supply projects, the only aspect on which there is positive consensus on project boundariesfrom the different guidance documents is that CO2 from the fossil fuels used to generate electricity/heatshould be included (which is where the bulk of the emissions come from). Different recommendations aremade for whether or not to include e.g. direct on-site emissions of CH4 and N2O and direct off-siteemissions arising from the transport of fuels to the project site. Similarly, recommendations on boundariesfor forestry projects do not agree whether soil carbon, roots and dead biomass (all of which can comprise asignificant proportion of total carbon stock on a project site) should be monitored.

Decision trees could be a useful means for project developers to decide whether or not to include particularemission sources in the project boundary and what appropriate indicators of project performance are. Usingsuch decision trees would increase comparability and consistency between projects, as well as facilitatingthe verification process (and therefore reducing its cost).

The frequency with which a project is monitored can influence the accuracy of emission mitigationestimates as well as the frequency with which project operators receive emission credits. However, there islittle guidance on appropriate monitoring frequencies for different project types, and little consistencybetween what guidance there is. Moreover, a distinction is needed in any guidance between the frequencywith which a particular parameter is monitored, and the frequency with which it is reported.

Less frequent monitoring and reporting (e.g. less than once per year) may be appropriate for some projecttypes, such as LULUCF projects, as changes in carbon stocks may be slow, gradual, and follow a well-known pattern. More frequent monitoring and reporting (e.g. monthly or weekly) may be appropriate forcertain parameters in other projects, such as energy production or efficiency projects, where theperformance is more variable throughout a year. Given that CERs are only issued after verification, andthat there will be a time delay between verification and issuance, it is likely that project operators may wishto verify projects annually - at least towards the end of a commitment period. However, because lessfrequent verification should lower total verification costs, project developers may also wish to have theoption of not verifying projects annually.

The costs of monitoring project performance will vary by type of project. For example, the time and costinvolved in calculating the GHG impacts of wind-powered electricity could be trivial (depending onproject boundary assumptions), but that for estimating carbon uptake by a forest is not. In general,monitoring costs will increase with the number of sources that need to be monitored; frequency of


37

monitoring and reporting; accuracy/confidence level that monitoring should achieve; and the complexity ofmonitoring methods.

Ideally, project monitoring should provide an accurate picture of the GHG impacts of a project. However,the credits generated by a project are calculated by comparing the project’s performance (which can bemeasured more or less accurately) with the project baseline (which is by definition a hypotheticalscenario). It may therefore be appropriate to make some trade-offs to ensure that monitoring costs are keptto a reasonable level, e.g. under 8% of a project’s total cost. Implications of possible actions to reducemonitoring costs are indicated in Table 8.

Table 8: Implications of possible actions to reduce monitoring costs

Effect of trade-off on:Type of trade-off to reducemonitoring costs

Numberof credits

Risk ofleakage

Accuracy of emissions/enhancement estimate

Verificationcosts

Narrower project boundary Lower orHigher

Higher Reduced for project aswhole but unchanged forsources actually monitored

Reduced

Reduced frequency (i.e.less than annual)monitoring

None (intheory)*

Low Reduced (for extrapolatedpoints), unchanged for otherpoints.

Increased**

Lower accuracy/confidence

None (intheory)*

Higher Reduced Increased

Simpler methods takinginto account conservativeassumptions andparameters

None (intheory)*

Unknown Reduced Reduced

In practice, it could be decided to reduce/discount the number of credits awarded to projects using simpler monitoringmethods, and/or lower accuracy/confidence levels and/or reduced monitoring frequency.

** Because it is more difficult to verify data where data series are not complete.

Trade-offs may sensibly be made at a different point for different project types. However, perhaps theeasiest means to address trade-offs is to use simple monitoring methods and conservative assumptions tomake sure that the number of credits generated by a project are not over-estimated. The implications ofdifferent trade-offs should be taken into account when making decisions on the ground rules that guide thedevelopment of monitoring guidance.


38

6. Annex A: Marrakech Accord provisions on the project cycle andproject monitoring

6.1 CDM project cycle

The Marrakech Accords lays out a number of specific steps that projects need to fulfil in order to becomeCDM projects and generate emissions credits. These steps are:

• Validation of project activity, which includes 30 days for stakeholders comment

• Registration of project activity, including a determination that the project is additional and that thebaseline is appropriate. The registration becomes final eight weeks after date of receipt of request forregistration.

• Monitoring projects throughout their crediting life, according to the monitoring plan. This is acondition for verification, certification and issuance of CERs.

• Verification of a project’s baseline, emissions and credits. An independent reviewer should undertakethis verification periodically, and ex post. Verification is also required for JI projects.

• Certification, which occurs on receipt of the verification report.

• Request for issuance of credits.

• Issuance of credits, which occurs after a 15 day delay.

6.2 Excerpts from Decision 17/CP.7

The following text is taken from Decision 17/CP.7 of the Marrakech Accords (UNFCCC 2001).

52. The project boundary shall encompass all anthropogenic emissions by sources of greenhouse gasesunder the control of the project participants that are significant and reasonably attributable to theCDM project activity.

Monitoring

53. Project participants shall include, as part of the project design document, a monitoring plan thatprovides for:

(a) The collection and archiving of all relevant data necessary for estimating or measuringanthropogenic emissions by sources of greenhouse gases occurring within the project boundaryduring the crediting period;

(b) The collection and archiving of all relevant data necessary for determining the baseline …

(c) The identification of all potential sources of, and the collection and archiving of data on, increasedanthropogenic emissions by sources of greenhouse gases outside the project boundary that aresignificant and reasonably attributable to the project activity during the crediting period;

….


39

(e) Quality assurance and control procedures for the monitoring process;

(f) Procedures for the periodic calculation of the reductions of anthropogenic emissions by sources bythe proposed CDM project activity, and for leakage effects;

(g) Documentation of all steps […] above.

…

54. For small-scale CDM project activities …. Project participants may use simplifiedmodalities and procedures for small-scale projects.

…

58. The implementation of the registered monitoring plan and its revisions, as applicable, shall be acondition for verification, certification and the issuance of CERs.

59. Subsequent to the monitoring and reporting of reductions in anthropogenic emissions, CERsresulting from a CDM project activity during a specified time period shall be calculated, applyingthe registered methodology, by subtracting the actual anthropogenic emissions by sources frombaseline emissions and adjusting for leakage.

Appendix B: Project design document

2. …. [a] project design document which shall include the following:

(h) Monitoring plan:(i) Identification of data needs and data quality with regard to accuracy, comparability,

completeness and validity;(ii) Methodologies to be used for data collection and monitoring including quality assurance and

quality control provisions for monitoring, collecting, and reporting;(iii) …

(j) Calculations:(i) Description of formulae used to calculate and estimate anthropogenic emissions by sources of

greenhouse gases of the CDM project activity within the project boundary;(ii) Description of formulae used to calculate and to project leakage, defined as: the net change of

anthropogenic emissions by sources of greenhouse gases which occurs outside the CDMproject activity boundary, and that is measurable and attributable to the CDM project activity;

(iii) …


40

Appendix C: Terms of reference for establishing guidelines on baselines and monitoringmethodologies

The executive board, drawing on experts … shall develop and recommend to the COP/MOP, inter alia:

(a) General guidance on methodologies relating to baselines and monitoring consistent with theprinciples set out in those modalities and procedures in order to:

…

(ii) Promote consistency, transparency and predictability;(iii) Provide rigour to ensure that net reductions in anthropogenic emissions are real and

measurable, and an accurate reflection of what has occurred within the project boundary…

(b) Specific guidance in the following areas:

…

(iii) Monitoring methodologies that provide an accurate measure of actual reductions inanthropogenic emissions as a result of the project activity, taking into account the need forconsistency and cost-effectiveness;

(iv) Decision trees and other methodological tools….…(vi) Determination of project boundaries including accounting for all greenhouse gases that

should be included as a part of the baseline, and monitoring.


41

7. References

Birner, Sabrina, personal communication (21.2.2002)

Bode, Jan-Willem, 2002, personal communication (22.2.2002)

Bode, Jan-Willem, R. H. Ravindranath, H. I. Somashekhar, 2001 - draft, Baseline Development forBioenergy Projects: Case Studies from South India

Brown, Sandra, Miguel Calmon, Matt Delaney, 1999, Carbon Inventory and Monitoring Plan for theGuaraqueçaba Climate Action Project, Brazil, Prepared by Winrock International for The NatureConservancy

Brown, Sandra, and Matt Delaney, 2001, Guaraqueçaba Climate Action Project: 2000 Findings and StatusReport, Prepared by Winrock International for The Nature Conservancy

Brown, Sandra, Ian Swingland, Robin Hanbury-Tenison, Ghillean Prance, Norman Myers, 2001, CarbonSinks for Abating Climate Change: Can they work?,http://www2.essex.ac.uk/ces/ResearchProgrammes/CESOccasionalPapers/ListOccPapers.htm

CDM Susac, 2002, How-to guide to Monitoring and Verification Protocol,http://cdmsusac.energyprojects.net/

DNV (Det Norske Veritas), 1999a, Illumex Draft Model Monitoring and Verification Protocol, Report 99-3288 prepared for the World Bank

DNV, 1999b, Illumex Verification Report, Report 99-3258 prepared for the World Bank

DNV, 2000, Slovakian-Norwegian JI projects verification report, prepared for the Norwegian Ministry ofEnvironment, report number 2000-3461

DNV, 2001a, Multi-Project Verification of Swedish AIJ Projects: Verification Results and Documentation,Report ER 10:2001 prepared for the Swedish National Energy Administration

DNV, 2001b, Multi-Project Verification of Swedish AIJ Projects: Methodology and Lessons Learned,Report ER 9:2001 prepared for the Swedish National Energy Administration

EcoSecurities Ltd, 2000, Financing and Financing Mechanisms for Joint Implementation (JI) Projects inthe Electricity Sector, November 2000, UK.EcoSecurities Ltd.

Ellis, Jane, 2001, Forestry Projects: Permanence, Credit Accounting and Lifetime, OECD and IEAInformation Paper COM/ENV/EPOC/IEA/SLT(2001)11,http://www.oecd.org/pdf/M00023000/M00023450.pdf

EZ (Ministry of Economic Affairs, the Netherlands), 2001, Operational Guidelines for Baseline Studies,Validation, Monitoring and Verification of Joint Implementation Projects,http://www.carboncredits.nl/

Factorag, 2001, Small-scale CDM Projects: Opportunities and Obstacles,http://www.factorag.ch/pdf%20files/SUT%20Small-scale%20CDM%20Projects_Vol1_5-11-01.pdf


42

Hamburg, Steven P, 2000, Simple rules for measuring changes in cecosystem carbon in forestry-offsetprojects, Mitigation and Adaptation Strategies for Global Change, 5: 25-37, 2000.

Heister, Johannes, personal communication, 5.4.2002

IGPO (International Greenhouse Partnerships Office), 2000, Workbook on Electricity and Heat Generationfrom Renewables, Prepared by Energy Strategies Pty Ltd and George Wilkenfield & Associates,http://www.isr.gov.au/resources/energy_greenhouse/igp

IPCC, 2000, IPCC Good Practice Guidance and Uncertainty Management in National Greenhouse GasInventories

IPCC/OECD/IEA, 1997, Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories

IPCC, 2000, Special Report on Land-Use, Land-Use Change and Forestry, http://www.ipcc.ch/

IPMVP, 2000, International Performance Measurement and Verification Protocol Volume I,http://www.ipmvp.org/info/downloads_2000.html

Kadyszewski, John, 2001, presentation at COP7 side event Forestry Projects: How to Credit and Monitor,http://www.iisd.ca/climate/cop7/enbots/nov5.html

Kartha, Sivan and Michael Lazarus, with Martina Bosi, 2002, Practical Baseline Recommendations forGreenhouse Gas Mitigation Projects in the Electric Power Sector, OECD and IEA InformationPaper

Loreti, Christopher, Scot Foster, Jane Obbagy, 2001, An Overview of Greenhouse Gas EmissionsVerification Issues, Prepared for the Pew Center on Global Climate Change,http://www.pewclimate.org/projects/emissions_verification.cfm

MacDicken, K.G., 1997, A Guide to Monitoring Carbon Storage in Forestry and Agroforestry Projects,Winrock International, http://www.winrock.org/

Martinot, Eric, 1998, Monitoring and Evaluation of Market Development in World Bank-GEF ClimateChange Projects: Framework and Guidelines, World Bank, paper number 066.

Martinsen, Thomas, Hans Jacob Mydske, Andy Yager, Milan Oravec, Ján Ilavský, 2002, Implementationof the AIJ projects Fuel Switch in Boilers in the Slovak Republic, Report prepared for the NorwegianPollution Control Authority

Nabuurs, G. J, A.J. Dolman, E. Verkaik, A. P. Whitmore, W. P. Daamen, O. Oenema, P. Kabat, G. M. J.Mohren, 1999, Resolving Issues on Terrestrial Biospheric Sinks in the Kyoto Protocol, DutchNational Research Programme on Global Air Pollution and Climate Change, Report no. 410 200 030(1999)

OECD/IEA, 2000, Emission Baselines: Estimating the Unknown, Paris

PCF (Prototype Carbon Fund), 2001a, Chile: Chacabuquito 26MW Run-of-River Hydro Project:Monitoring and Verification Protocol,http://www.prototypecarbonfund.org/router.cfm?show=Discuss.cfm


43

PCF (Prototype Carbon Fund), 2001b, Uganda: West Nile Electrification Project Monitoring andVerification Protocol, http://www.prototypecarbonfund.org/

PCF, 2000, Learning from the Implementation of the Prototype Carbon Fund, PCF Occasional PapersNumber 1

Powell, M. 1999, Effect of Inventory Precision and Variance on the Estimated Number of Sample Plotsand Inventory Variable Cost: The Noel Kempff Mercado Climate Action Project. Paper prepared forthe World Bank, Winrock International, 38 Winrock Dr., Morrilton, Arkansas 72110, USA.

Ravindranath, N. H., and P. R. Bhat, 1997, Monitoring of carbon abatement in forestry projects - Casestudy of Western Ghat project, Mitigation and Adaptation Strategies for Global Change, 2, 217-230

Richard Tipper, William McGhee, Jane Ellis, Ben H. de Jong, Augustine Hellier, 2001, An Initial View onMethodologies for Emission Baselines: Forestry Case Study (internal OECD/IEA consultationpaper)

UNEP/OECD/IEA, 2001, Baseline methodologies: Possibilities for Standardised Baselines for JI and theCDM, Document COM/ENV/EPOC/IEA/SLT(2001)15http://www.oecd.org/pdf/M00007000/M00007696.pdf

UNFCCC, 2001, The Marrakech Accords and the Marrakech Declaration

UNFCCC, 2001b, Activities Implemented Jointly Under the AIJ Pilot Phase: Balvi (II), District heatingrenovation project; http://unfccc.int/program/aij/aijact01/Sweden/lvaswe-04-01.html

UNFCCC, 2001c, Activities Implemented Jointly Under the AIJ Pilot Phase: Reduction of AtmosphericPollution through modernisation of HEAT Supply system in the Town of ByCzyna,http://unfccc.int/program/aij/aijact01/polnld-01-01.html

UNFCCC, 2000, Activities Implemented Jointly Under the AIJ Pilot Phase: Improving Thermal PowerPlant Efficiency, http://unfccc.int/program/aij/aijact01/Jorduitfrcn-01.html

UNFCCC, 1998, Activities Implemented Jointly Under the AIJ Pilot Phase: Klinki forestry project

USDOE, 1994, Sector-specific Issues and Reporting Methodologies Supporting the General Guidelines forthe Voluntary Reporting of Greenhouse Gases under Section 1605(b) of the Energy Policy Act of1992: Forestry Sector, http://www.eia.doe.gov/oiaf/1605/guidelins.html

Vine, Edward and Jayant Sathaye, 1999, Guidelines for the Monitoring, Evaluation, Reporting,Verification, and Certification of Energy-Efficiency Projects for Climate Change Mitigation, ReportLBNL-41543

Vine, Edward, Jayant Sathaye and Willy Makundi, 1999, Guidelines for the Monitoring, Evaluation,Reporting, Verification, and Certification of Forestry Projects for Climate Change Mitigation,Report LBNL-41877

WB (World Bank), 1998, Greenhouse Gas Assessment Handbook,http://wbln0018.worldbank.org/essd/kb.nsf/ab34a7716339450985256666007c9ca9/901594d74185a58885256761007cbefd


44

WBCSD/WRI, 2001, The Greenhouse Gas Protocol: a corporate accounting and reporting standard,http://www.wbcsd.org

WGB (Working Group on Baseline for CDM/JI Project), 2001, Technical Procedures for CDM/JI Projectsat the Planning Stage - Interim Report to the Ministry of the Environment, Government of Japan

Willems, Stéphane, 2000, Key Features of Domestic Monitoring Systems under the Kyoto Protocol,OECD and IEA Information Paper COM/ENV/EPOC/IEA/SLT/(2000)4,http://www.oecd.org/env/cc/


45

8. Glossary

AIJ Activities Implemented Jointly

CDM Clean Development Mechanism

CERs Certified Emission Reductions

COP Conference of the Parties

EB The Executive Board of the Clean Development Mechanism

ERUs Emission Reduction Units

GHG Greenhouse gas

GWP Global warming potential

IPMVP International Performance Measurement and VerificationProtocol

JI Joint Implementation

LULUCF Land-use, Land-use change and forestry

Necromass Dead biomass (lying or standing)

OE Operational Entity

PCF Prototype Carbon Fund

QA/QC Quality assurance and quality control procedures.

Spillover Spillover refers to positive side-effects of the project (i.e.increased emission reductions or sequestration) that are notincluded in the project boundary and that do not generateemission credits.

Documents

DEVELOPING GUIDANCE ON MONITORING AND PROJECT … · developing guidance on monitoring and project boundaries for greenhouse gas projects information paper. com/env/epoc/iea/slt(2002)2